A Model for a Smallpox-Vaccination Policy

Published in: The New England Journal of Medicine, v. 348, no. 5, Jan 30, 2003, p. 416-425

Posted on RAND.org on January 01, 2003

by Samuel A. Bozzette, Rob Boer, Vibha Bhatnagar, Jennifer Brower, Emmett B. Keeler, Sally C. Morton, Michael A. Stoto

Read More

Access further information on this document at content.nejm.org

This article was published outside of RAND. The full text of the article can be found at the link above.

The new reality of biologic terrorism and warfare has ignited a debate about whether to reintroduce smallpox vaccination. The authors developed scenarios of smallpox attacks and built a stochastic model of outcomes under various control policies. They conducted a systematic literature review and estimated model parameters on the basis of European and North American outbreaks since World War II. They also assessed the trade-offs between vaccine-related harms and benefits. Nations or terrorists possessing a smallpox weapon could feasibly mount attacks that vary with respect to tactical complexity and target size, and patterns of spread can be expected to vary according to whether index patients are hospitalized early. For acceptable results, vaccination of contacts must be accompanied by effective isolation. Vaccination of contacts plus isolation is expected to result in 7 deaths (from vaccine or smallpox) in a scenario involving the release of variola virus from a laboratory, 19 deaths in a human-vector scenario, 300 deaths in a building-attack scenario, 2735 deaths in a scenario involving a low-impact airport attack, and 54,728 deaths in a scenario involving a high-impact airport attack. Immediate vaccination of the public in an attacked region would provide little additional benefit. Prior vaccination of health care workers, who would be disproportionately affected, would save lives in large local or national attacks but would cause 25 deaths nationally. Prior vaccination of health care workers and the public would save lives in a national attack but would cause 482 deaths nationally. The expected net benefits of vaccination depend on the assessed probability of an attack. Prior vaccination of health care workers would be expected to save lives if the probability of a building attack exceeded 0.22 or if the probability of a high-impact airport attack exceeded 0.002. The probability would have to be much higher to make vaccination of the public life-saving. Conclusions. The analysis favors prior vaccination of health care workers unless the likelihood of any attack is very low, but it favors vaccination of the public only if the likelihood of a national attack or of multiple attacks is high.

This report is part of the RAND Corporation external publication series. Many RAND studies are published in peer-reviewed scholarly journals, as chapters in commercial books, or as documents published by other organizations.

The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.